Will Do Better - Fabrizio Cariani

Will do better ∗
Fabrizio Cariani
Northwestern University
Paolo Santorio
University of Leeds
Abstract Statements about the future are central in everyday conversation and
reasoning. How should we understand their meaning? The received view among
philosophers treats will as a tense: in “Cynthia will pass her exam”, will shifts the
reference time forward. Linguists, however, have produced substantial evidence
for the view that will is a modal, on a par with must and would. The different
accounts are designed to satisfy different theoretical constraints, apparently
pulling in opposite directions. We show that these constraints are jointly satisfied
by a novel modal account of will. On this account, will is a modal but doesn’t
work as a quantifier over worlds. Rather, the meaning of will involves a selection
function similar to the one used by Stalnaker in his semantics for conditionals.
The resulting theory yields a plausible semantics and logic for will and vindicates
our intuitive views about the attitudes that rational agents should have towards
future-directed contents.
[Draft of June 9, 2015]
Our topic is the semantics for statements about the future in English. In particular,
we focus on sentences involving the English auxiliary will, such as:
Cynthia will pass her exam.
Sentences like (1) are uniquely interesting. An account of their meaning faces
challenges from a number of philosophical domains: semantics, epistemology,
and metaphysics.
The semantic challenge is generated by a tension in the linguistic behavior
of will. On the one hand, will has the characteristic marks of a modal operator.
On the other, will fails to display the standard scope interactions of modals. For
example, unlike must or might, will commutes freely with negation. That is,
“It will be the case that it doesn’t rain” and “It is not the case that it will rain”
have the same truth conditions, despite the difference in relative scope between
will and negation.
∗ For conversations and exchanges we thank [omitted for blind review]
The epistemological challenge comes from the role of will-statements in everyday thinking and deliberation. We are often uncertain about the propositions
expressed by will-claims, and at least sometimes this uncertainty seems rational.
An adequate account of will should assign will-sentences contents towards which
we do, and rationally may, have attitudes of this sort.
The metaphysical challenge comes from considerations about the open future.
Past facts are settled, while at least some of the facts about the future seem not
to be. Both the claim that the future is open and the nature of the relationship
between metaphysics and semantics of the future are disputed. But the following
seems uncontroversial: whatever the truth about the metaphysics, our semantic
theory should avoid ascribing widespread error to ordinary speakers.
The existing literature exhibits a split between philosophers and semanticists.
Philosophers generally treat will as a tense, i.e. an operator whose semantic
function is to shift the time of evaluation of a clause. This view is well-positioned
to accommodate the epistemological challenge, and can be developed so as to
meet the metaphysical challenge. But it is problematic from a linguistic point
of view. Linguists (in particular, semanticists) treat will as a modal—i.e., as an
expression that manipulates a world parameter. Though typical modal accounts
of will are well-positioned to accommodate the linguistic challenge, they flounder
in the face of the epistemological challenge.
We aim to propose a new theory of will that improves on all existing accounts
and draws together elements from philosophy and linguistics. Here is a sketch.
Following semanticists, we hold that will is a modal. But will differs from standard
modals like must or may, which work as quantifiers. The best analogy for will is
the selection function meaning that Robert Stalnaker uses in his semantics of
conditionals. That is, will selects a unique world among the ones that are included
within a domain of quantification: roughly, this is ‘the world instantiating the
one actual complete course of history’, among the ones that are compatible with
history up to now. The approximate truth conditions of (1) are:
In the actual complete course of history, Cynthia passes her exam.
Hence our semantics presupposes that there is a ‘unique’ actual course of history.
At the same time, it might be indeterminate which possible world instantiates
the actual course of history. As a result, it might be indeterminate which world
will selects, and will-statements may have indeterminate truth value. The resulting view (i) yields a plausible semantics and logic for will, (ii) generates contents
for will-statements towards which we can be rationally uncertain, and (iii) makes
room for (though doesn’t require) the metaphysical claim that the future is open.
Will do better
We proceed as follows. In §2, we outline some plausible constraints for a
semantics for will. In §3, we give an informal overview of the account, which is
stated in full in §5 (§4 spells out our metaphysical assumptions). In §6 we explain
how to define a notion of truth that makes room for indeterminacy. Finally, we
check that our account yields the desired logical and epistemological predictions
Semantics for the future: three constraints
Any plausible account of will, we maintain, ought to respect three constraints.
The modal character of will
Our first constraint is that will has a modal meaning. By this we mean that
will manipulates a possible world parameter,1 similarly to modal auxiliaries like
must or might. For example, here is a toy modal meaning for will:
ðwill Añ is true at w and t iff for all worlds w0 in a relevant domain
in w, A is true at w0 and at some t 0 ≥ t
The modal view contrasts with a temporal view, on which will manipulates
exclusively a time parameter and no world parameter. Again, for illustration,
here is a toy temporal view:
ðwill Añ is true at w and t iff for some t 0 ≥ t, A is true at t 0 and w
To be clear: we understand the modal view as compatible with the claim that,
in addition to the world parameter, will manipulates a time parameter. What
distinguishes modal from nonmodal analyses is whether will manipulates a world
parameter at all.
The linguistics literature has provided three pieces of evidence for the modal
view. Taken together, they seem to us decisive. The first piece of evidence is
morphological. On one widely accepted account (Abusch 1997, 1998, Condoravdi 2002, Kaufmann 2005), will shares morphology with the modal would. In
particular, will and would have in common a modal morpheme, often represented
as ‘WOLL’: will is PRESENT + WOLL; would is PAST + WOLL.
The assumption of common morphology allows us to explain otherwise
puzzling semantic facts. For example, it explains why we can replace will with
1 It doesn’t matter to our account whether this world parameter is assumed to be represented in
the object-language via world variables, or in the metalanguage via an index coordinate. When
stating our semantics, we choose the latter option.
would in indirect reports of past utterances of will-sentences. If, on Tuesday,
Harriet says “I will come to work tomorrow”, then on Wednesday we would
report Harriet’s utterance by saying “Harriet said she would come to work today”.
The second piece of evidence for the modal view of is that will (on a par with
other expressions that normally induce future reference, like going to) may have
epistemic readings (Palmer 1987, Enç 1996). These readings generally require a
stative predicate (like be) in the prejacent.2 Here is an example:
John will be in London by now.
Notice two facts about (3). First, as is made clear by the modifier by now, the
prejacent of (3) has its reference time in the present. Hence, at the time of
utterance, it is settled whether John is in London. Second, in (3), will works
as a marker of evidentiality: roughly, it signals that the speaker is inferring
John’s location on the basis of a body of evidence. To see this, notice that (3) is
infelicitous if uttered by someone who is looking directly at John, even if both
are indeed in London. Both these facts are hard to explain on a purely temporal
view. By contrast, views according to which our language of uncertainty and
prediction are both modal in nature seem ideally placed to account for them.
Third, as Peter Klecha (2013) has recently argued, will allows for modal
subordination. Roughly, modal subordination is the phenomenon whereby, in
discourses containing several modals, earlier modals may be used to restrict the
domain of later modals (Roberts 1989). As an example, consider the following
Jane might come to the party. Sally would come too.
The occurrence of would in the second sentence is naturally understood as
restricted to worlds where Jane does come to the party. A natural explanation for
this is that its domain of quantification is somehow anaphoric to the worlds that
witness the might-claim in (4). Now, Klecha points out that, similarly to what
happens with would and other modals, will-sentences can inherit restrictions
from previous modal elements of the discourse.
If the supplies arrive tomorrow, it will be late in the day. They will contain
three boxes of cereal.
This makes these sentences pattern together with modals like might and must,
and differently from tenses, like the past tense:
2 Following common (and medieval) usage, we use the term prejacent to denote the clause that
will takes scope over.
Will do better
If the supplies arrive tomorrow, it might (/must) be late in the day.
They might (/must) contain three boxes of cereal.
b. #If the supplies arrived yesterday, it was late in the day. They contained
three boxes of cereals.
Our second constraint is that will is scopeless with respect to an important class
of other linguistic items. By this we mean that changes in the relative syntactic
scope between will and these other items don’t make a difference to the truthconditions of will-sentences. This is a remarkable feature of will, and one that is
not generally shared by modal expressions. For present purposes, it is sufficient
to observe scopelessness with respect to negative items.3 Consider the following
It will not rain.
It is not the case that it will rain.
(7-a) and (7-b) are truth-conditionally equivalent. The situation is similar with
different prejacents, and when clauses like (7-a) and (7-b) appear embedded
in other environments. In short, will appears to commute freely with ordinary
English negation. This observation is strengthened by considering items that
lexicalize negation, such as doubt (which, following common assumptions, we
understand as believe that not) and fail: (8-a-b) are truth-conditionally equivalent.
I doubt that Sam will pass his logic exam.
I believe that Sam will fail his logic exam.
For a comparison with an auxiliary that is not scopeless, consider minimal variants
of (8-a-b) which involve a deontic modal. Suppose we’re talking about the
obligations that Sam must fulfill in order to stay enrolled in his degree. It is clear
that (9-a) is not truth-conditionally equivalent to (9-b).
I doubt that [in order to graduate] Sam must pass his logic exam.
I believe that [in order to graduate] Sam must fail his logic exam.
The lack of scope interactions with negation immediately yields an interesting
logical constraint:
3 Perhaps the scopelessness of will extends further. One further category of items with respect to
which will might be scopeless is comparative expressions (see the discussion of comparatives
and conditionals in Korzukhin 2014). For reasons of space, we limit ourselves to considering
Will Excluded Middle (preliminary take): ðwill A ∨ will not Añ
is a logical truth
For now, we informally gloss ‘logically true’ as ‘true whenever uttered’. In §7, we
derive the validity of this schema, given our semantics and two standard formal
concepts of consequence.
The cognitive role of future statements
Future-directed statements play an important role in our cognitive economy. It
is a platitude that ordinary agents are uncertain about the future. Assuming
that credences attach to propositions, it seems natural to understand ‘being
uncertain about the future’ as ‘having a non-extreme degrees of belief towards
the propositions that are expressed by will-claims’. Moreover, at least in some
cases, these non-extreme degrees of belief seem also rational.
Sports Fan: Suppose that Cynthia comes to work each day wearing
a Warriors cap, a Giants cap or no cap, depending on a random
draw (with each option having equal probability). What degree
of belief should you assign to the proposition that tomorrow she’ll
wear a Warriors cap?
Presumably "1/3" is a rationally permissible answer. In some theoretical settings,
it may even be required: if some version of Lewis’s (1986c) principal principle is
a requirement of rationality and all of your evidence is of the admissible variety,
"1/3" would appear to be the only rational answer. Similarly, it seems that the
fair odds for a bet on the truth of that proposition are 1 to 2.
These seem like truisms, yet they are surprisingly hard to vindicate on existing
modal accounts. For illustration, consider the toy modal semantics mentioned
in §2.1. On that semantics, will is a universal quantifier over a domain of
worlds. On a natural construal of the domain of quantification of will, on this
semantics ðwill Añ says that A is true at all futures that are ‘open’ from the point
of utterance (this semantic account captures what Prior (1967) calls the ‘Peircean
future tense’). The problem for this semantics is that if Warriors-cap futures
and Giants-cap futures are both possible, you should have zero credence in the
propositions expressed by each of (10) and (11).
Cynthia will wear a Warriors cap tomorrow.
Cynthia will wear a Giants cap tomorrow.
Will do better
Any theory that treats will as a universal quantifier faces this problem.4
Taking stock
It is difficult to satisfy all three constraints. For one thing, the first two seem
to be in direct tension with each other. If will is a modal, as the first constraint
requires, we expect it to have nontrivial scope interactions, in violation of the
second constraint. And indeed, basic modal analyses of will predict that switching
the relative scope of will and negation does have truth conditional effects. For
illustration: on Kaufmann’s (2005) account, will is a universal quantifier over
(roughly) the worlds realizing the most likely courses of future history. On this
theory, by switching around will and not, we get the two nonequivalent readings:
will > not: all most likely futures do not satisfy the prejacent.
not > will: not all the most likely futures satisfy the prejacent.
The linguistics literature offers attempts to reconcile the first two desiderata.
A prominent example is the modal analysis of Copley (2009). Building on work by
von Fintel on generics (1997), Copley claims that will-sentences presuppose that
their domain is homogeneous with respect to the prejacent. For an occurrence
of will in a sentence of the form ðwill Añ to have a denotation, its domain
must contain only A-worlds or only ¬A-worlds. We have concerns about the
stipulative character of this proposal. But we can set them aside, because the
theory runs into a more basic problem: it fails to address our third constraint. The
propositions that this theory delivers are not propositions that we can plausibly
have nonextreme credences in. To see this, consider again (1):
Cynthia will pass her exam.
Simplifying, the basic truth conditions of (1) on both Kaufmann’s and Copley’s
views are that it is settled that Cynthia will pass the exam. Now, suppose that
you believe that the future is genuinely open with regard to Cynthia passing the
exam: it is unsettled whether she will. Then your credence in (1) must be zero:
you are certain that the truth condition of (1) fails to hold.5
4 The problem is analogous to one that has recently received attention in the counterfactuals
literature (see Edgington 2008, Moss 2014, Schulz 2014).
5 We do not claim that it is impossible for Copley to extend her view to account for these facts
about credences. It might well be possible to replicate in her framework some of the moves we
make in §8. We just claim here that Copley’s framework cannot be taken as is to account for the
cognitive role of future statement. Moreover, the solution to the cognitive problem we will devise
is more straightforwardly applied to the semantics we go on to construct.
Let us now peek quickly at the philosophical literature. By far the most
popular view among philosophers is what Prior calls ‘Ockhamist semantics’.
Ockhamists don’t ascribe any modal character to will. For them, “It will rain”
is true if and only if there is a future moment at which it rains (context might
further narrow down the interval during which this moment is situated). The
obvious problem with Ockhamism is its inability to satisfy our first desideratum.
The Ockhamist has no story about the relationship between will and would, about
predictive uses of will, or about modal subordination.6
One final option is to claim that will is ambiguous between a modal and a
nonmodal meaning. The modal meaning explains why will seems to satisfy the
first desideratum. The nonmodal meaning explains why it seems to satisfy the
second and third desiderata. We won’t attempt a full refutation of the ambiguity
option; but we notice that it has two major disadvantages. First, it systematically
overgenerates. For example, it predicts a true and a false reading for:
The probability that Cynthia will wear a Warriors cap is 0.
(12) is true on the modal meaning and false on the nonmodal meaning, so we
should be able to hear it as true. Perhaps there are maneuvers to be made to
block this reading, but we leave it to the ambiguity theorist to explain what they
are. Second, an account of will that does not exploit ambiguity seems obviously
preferable on the usual grounds of simplicity and theoretical unity. So, by giving
an account of will that satisfies all desiderata we provide an indirect argument
against the ambiguity view.
Overview: selecting the future
We present our full account of will in the next few sections, but it is helpful to
illustrate the central ideas without the formalism. We start by adopting some
(but not all) of the insights associated with branching time frameworks. At every
moment in time, we suppose, there are multiple possible histories that fully
coincide with respect to the past and diverge with respect to the future. As time
6 It might appear that classical supervaluationism is an exception to this pattern. Classical supervaluationists have a two-step approach: first, define truth relative to a single possible future in
the same way as the Ockhamist semantics. Next, say that A is true simpliciter just in case it is true
at the time of utterance in every objectively possible future. However, as Cariani (unpublished)
points out, supervaluationism is not a modal theory of will. Though it has a modal element, it
is not distinctive of the lexical entry for will, but in the ‘global’ definition of truth at a context.
For this reason, classical supervaluationism is unable to account for the evidence for a modal
treatment of will we summarized in §2.1. This does not entail that supervaluationist technology
is not useful. Indeed, we will avail ourselves to a non-classical form of supervaluationism in §6.
Will do better
passes, histories that had previously coincided up to a point part ways, ‘making
true’ different courses of events. In diagram form:
This picture is often combined with substantial metaphysical claims about the
nature of possible worlds and the indeterminacy of the future. Importantly, our
account is neutral between all the relevant options (we clarify this in §4). All
we need is that, given any world w and time t, we can determine the historical
alternatives to w at t. Here is how we do it:
Two worlds w and v are historical alternatives at t iff w and v
match perfectly in their history (i.e., iff they match perfectly in
matters of particular fact) up to t.
Note that our definition of historical alternatives involves no reference to a
notion of openness. This is key to our later vindication (§4) of the claim that our
semantics is neutral about the open future hypothesis.
Now, consider again (10), repeated here:
Cynthia will wear a Warriors cap tomorrow
As a first step, we assume that will is a modal. Like all modals in natural language,
it is interpreted against a background set of possibilities. Following Kratzer’s
terminology (1977, 1981a, 1991b), we call this set the modal base. For the
particular case of will, the modal base in a given context is the set of historical
alternatives to the world of the context, at the time of the context. For example,
in the scenario we described, the modal base of (10) includes worlds where
Cynthia wears a Warriors cap, worlds where she wears a Giants cap, and worlds
where she wears no cap. In diagram form:
Warriors cap
Giants cap
no cap
Figure 1
Standardly, the modal base is the domain of quantification of the modal. But,
as we anticipated, our account is not quantificational. Instead, we propose that
will singles out one world within the modal base, and evaluates the prejacent at
that world. Intuitively, the selected world represents the ‘way things will actually
be’—in other words, the historical alternative that will actually be realized. So,
(10) is true just in case, in the selected world, Cynthia wears a Warriors cap
Warriors cap
Giants cap
no cap
Figure 2
The explanation for the scopelessness of will (and consequently the validation
of will-excluded middle) flows immediately from the fact that the prejacent is
always evaluated relative to a single world (see §7).
Our semantics for will presupposes that, at the time of utterance, there is a
unique, fully specified way things will actually be (in the jargon introduced by
Belnap & Green 1994, this is the assumption that there is a ‘thin red line’ that
marks the complete course of actual history). This assumption is controversial.
Theorists in the branching time tradition object that, in the context of futuredirected discourse, we have no right to speak of ‘the way things will actually be’,
or of ‘the actual world’. On the one hand, it might be that the future is open—that
there is no fact of the matter, at the current time, about what way things will turn
out. On the other, even if the future is not open, it is not clear that a semantics
for natural language can legitimately presuppose a metaphysical claim of this
Will do better
Even if one agrees with these concerns, we don’t think that the compositional
semantics for will needs to be changed. We distinguish what information is
needed by the compositional semantics from what information the world is able
to supply. We assume that the compositional semantics requires as input a unique
world of utterance. Like all parameters used in semantic computations, the value
of this parameter is supplied by the context. At the same time, we leave it open
that it may be indeterminate what context the utterance takes place in, and hence
what world is supplied to the semantics.
For illustration, consider again (10). Perhaps, at the time of utterance, it is
indeterminate whether the actual world is a Warriors-cap-world, a Giants-capworld, or a no-cap-world. If so, it is indeterminate which context the utterance of
(10) takes place in. The context might be the context of Figure 2, or it might be
a context in which some other world (for example, v, as in Figure 3) is selected.
Giants cap
Warriors cap
no cap
Figure 3
Let us highlight an important point. We grant that it may be indeterminate what
world an utterance takes place in; moreover, we said that the modal base of will is
determined as a function of the world and the time of the context. But it doesn’t
follow that the modal base of will is indeterminate. The reason is that, given the
way that we have defined historical alternatives, all worlds that are candidates for
being the world of the context have the same historical alternatives. So we will
be able to speak of the modal base of will in a context, even if it is indeterminate
what context the utterance takes place in.
The next few sections implement the plan we just sketched. §4 specifies some
metaphysical background. §5 presents our compositional semantics, including an
analysis of will-conditionals. §6 elaborates on our treatment of indeterminacy; §7
shows how our account yields the logical and linguistic predictions we identified;
§8 shows how our account yields appropriate predictions about cognitive role.
Metaphysical background
Our account is neutral on a number of metaphysical issues connected to branching. In this brief section we explain how.
First, supporters of branching time often claim that possible worlds literally
share temporal initial segments (see, e.g., Thomason 1970, Belnap & Green 1994,
Belnap et al. 2001). The point at which two worlds branch is the point at which
the initial segment ends. By contrast, opponents of branching argue that worlds
with identical histories up to a point are qualitatively identical, but still have no
parts in common (see, e.g., Lewis 1986a). We understand the claim that two
worlds w1 and w2 are historical alternatives at a time t as the weak claim that
there is perfect match between matters of particular fact between w1 and w2 up
to t. This is compatible both with genuine overlap and with mere qualitative
Second, the branching framework is often associated with the claim that the
future is ‘open’. The relevant contrast here is with past events, which are taken
to be fixed in a way in which the future is not. There are a number of ways to
explain the relevant concept of openness. Following Barnes & Cameron (2009),
we choose one that is noncommittal between different metaphysical theses:
Openness: (at least some) contingent facts about how things will
be are presently unsettled (in a suitably objective sense).
Some writers (for example, Belnap & Green 1994) adopt Openness as the starting
point of the enterprise of giving a semantics for will. Others (like MacFarlane
2003, 2008, 2014) take it as a methodological desideratum that a semantics
for the future should not decide between different metaphysical options about
We are not committed to any of these claims. Unlike Belnap and Green, we
don’t assume Openness. Our apparatus is compatible with both Openness and its
denial. Unlike MacFarlane, we don’t endorse the neutrality of the semantics as a
methodological constraint. As it happens, however, our semantics for will does
turn out to be metaphysically neutral about Openness—in the sense that both
the defender and the opponent of Openness are able to use it. The reason is that,
as we pointed out in §2, there is strong independent reason to adopt a modal
semantics for will. This kind of modal semantics lends itself well to the sort of
neutrality about Openness we just characterized. But we do not claim that the
metaphysical neutrality is, in itself, a reason to accept our account.
Will do better
Let us start by introducing some notation. We use sans-serif letters (‘A’, ‘B’, etc.)
as metalinguistic variables over sentences; and boldface letters (‘A’, ‘B’, etc.)
as metalinguistic variables over sets of worlds. We use ‘propositions’ and ‘sets
of worlds’ interchangeably, but everything we say is meant to be compatible
with theories according to which propositions merely determine sets of possible
worlds, without being identical to them.
As is standard in semantics, we define an interpretation function of the form,
¹·º par amet ers,g
Such a function assigns compositional truth-values to sentences relative to a
series of parameters and an assignment function, conventionally denoted by
‘g’ (the latter is just a function that assigns objects to syntactic indices, and is
needed to handle variables). Different theories employ different parameters.
The interpretation function is also relativized to a context, but to remove clutter
we avoid explicit mention of the context unless strictly needed.
We also make some specific assumptions about will. First, will is a sentential
operator, i.e. an operator that takes a full clause as argument. This is a simplification, but one that is harmless given our purposes. Second, as we flagged
in §3, will takes as argument a modal base, i.e. a set of worlds that are used
for the interpretation of the modal.7 In particular, the modal base of will is
the set of historical alternatives to the world of the context. Syntactically, we
assume that modal bases are the semantic values of covert pronouns that work
as arguments of modals. We represent these pronouns as ‘ f i ’, and their values,
a set of worlds, as ‘Fi ’. For shorthand, we generally represent modal bases in
LFs just as a subscript of modals; hence we write ‘will f ’ rather than the more
extended ‘will [ f ]’.
Semantics for will
Our semantics for will is based on an extended analogy with Stalnaker’s (1968)
semantics for conditionals.8
7 This understanding of modal bases is slightly simplistic. Modal bases are officially functions from
worlds to propositions (see von Fintel & Heim 2011 for discussion).
8 Schulz (2014) has recently defended an interesting variant of Stalnaker’s semantics. Roughly,
conditionals quantify over a set of worlds, but they also select (via a choice function) an arbitrary
world within that set. We lack space for a full comparison here. Let us just state without argument
Similarly to what happens on Stalnaker’s semantics, we assume that the
interpretation of will involves appeal to a selection function, denoted by ‘s’. A
selection function maps a pair of a world w and a proposition A to a ‘selected’
world w0 . Intuitively, s selects the world w0 that is ‘closest’ to the starting world
w while at the same time verifying proposition A. For the case of conditionals,
and counterfactuals in particular, there is much literature on how exactly the
metric of closeness should be construed.9 We don’t need to settle these issues
here. We can adopt any of the metrics that have been proposed for counterfactual
Selection functions are characterized by two important constraints:
Inclusion: if A is non-empty, s(w,A) ∈ A
Centering: if w ∈ A, s(w,A) = w.
Inclusion says that the world selected must verify the input proposition (provided
that some world does verify the input proposition). Centering says that, if the
input world verifies the input proposition, then the world selected is the input
world itself. Inclusion and centering are the only constraints we impose on
selection functions, which can then be defined as follows:11
A function s : W × P (W ) 7→ W is a selection function iff
i. if A is non-empty, s(w,A) ∈ A, and
ii. if w ∈ A, then s(w,A) = w.
At this point, we’re ready to state the meaning of will. We assume that interpretation is relativized to three parameters:12 a world of evaluation w, a selection
function s, and an assignment g.
¹will f Aºw,s,g = 1 iff ¹Aºs(w,g( f )),s,g = 1
or development that, on the most natural implementation, building the analogy with Schulz’s
system rather than Stalnaker’s would require us to consider indeterminacy at the level of the
compositional semantics, which we are reluctant to do.
9 For some sample proposals concerning the metric of closeness for counterfactuals, see Lewis
(1979a), Kratzer (1981b), Hiddleston (2005).
10 As will be evident soon, the choice between different metrics only matters for will-conditionals.
All we need to settle the selected world in all other cases is merely the Centering condition (see
11 Stalnaker imposes some extra constraints on selection functions. We leave it open whether these
extra constraints should apply to will; nothing hinges on these for our purposes.
12 Recall: to avoid clutter, we omit the context parameter.
Will do better
To simplify the notation, we will just write F instead of g( f ) throughout the
Let us make some comments about this compositional semantics.
First, the basic effect of will is to shift the world at which its prejacent is
evaluated. This feature is shared with standard semantic accounts of modals
in natural language, like must. The difference is that modals usually introduce
quantification over the world of evaluation parameter, while will replaces the
world of evaluation for another one picked via the selection function.
Second, the entry in (13) does not reflect any temporal shift. It is easy to
introduce temporal shift, letting will quantify existentially over times (accordingly,
interpretation is relativized to an extra parameter for times).
¹will f Aºw,t,s,g = 1 iff ∃t 0 ≥ t,¹Aºs(w,F),t ,s,g = 1
This said, throughout the discussion we just stick to the entry in (13). This is
mostly for simplicity. The central innovation we introduce is the appeal to selection functions. Other elements of the meaning of will can stay in the background.
Moreover, there are reasons to think that a full-blown semantics for will exploits
time in a way that is more complex than simple existential quantification over
temporal instants.13 So the account in (14) would need update and clarification
Third, this semantics has an interesting consequence for unembedded occurrences of will: as it turns out, will is semantically vacuous with respect to
the modal parameter. Recall that the modal base of will defaults to the set of
historical alternatives to the world of the context. Furthermore, the initial world
of evaluation defaults to the world in which the utterance takes place.14 Hence,
when will is unembedded, the world that works as the input to the selection function is a member of the modal base. In this situation, the centering assumption
entails that the world returned by the selection function is always the world of
evaluation itself. Thus, in its unembedded occurrences, will merely ‘overwrites’
the world of evaluation parameter with itself.
¹will f Aºw,s,g = 1 iff ¹Aºw,s,g = 1
13 In particular, as argued by Abusch (1998) (see also Condoravdi 2002), it seems that our best
semantics for tense should quantify over intervals rather than instants. In the context of this
theory, the semantic effect of will (as well as of other modals) would be not to shift the time of
evaluation, but rather to extend forward the time interval at which the prejacent is evaluated
(for a proposal in this vein, see also Kaufmann 2005).
14 This is via the definition of truth at a context, which fixes the value of w to the world of the
context. See section 6 for details.
Thus, when will occurs unembedded, our semantics effectively collapses on the
simple Ockhamist semantics, which treated will as a mere tense.
Why bother, then, with the complexities of our selection function semantics?
There are many good reasons. They mainly relate to the fact that, on our
account, will has a modal base. This opens up the possibility of accounting for
will-conditionals (adopting the popular assumption that will-clauses function as
restrictors), modal subordination (via anaphoric links between the modal bases
of the different modals), and epistemic readings (by assuming that the modal
base can have different flavors). The selection function account also allows us
for a vindication the will-would connection.
Giving a full-fledged account of all these phenomena would take too long.
But below we give a brief sketch. Even from these quick remarks, it should be
clear that our account provides the tools for vindicating in full all the marks of
the modal character of will.
Conditionals and modal subordination
A selection function semantics for will allows for a natural account of willconditionals. This will also allow us to account for modal subordination.
Following a longstanding tradition (see e.g. Lewis 1975, Kratzer 1991a, 2012)
we assume that the function of if -clauses is to restrict modal bases—to rule out
of the modal base the worlds that are incompatible with them (this effect is
modeled by intersecting the modal base with the set of worlds individuated by
the if -clause).
There are many ways to implement this semantic effect. Building on work on
quantifier domain restriction by Kai von Fintel (1994), we choose a simple one
that dovetails well with our assumption that the object language syntax contains
a variable referring to the modal base. Nothing hinges on this particular choice
of implementation.
We assume that if -clauses work as assignment shifters (similarly to lambdabinders in a system in the style of Heim & Kratzer 1998). At a syntactic level,
if -clauses are coindexed with the relevant modal base variable. For example, the
LF of (16) is in (17):
If John goes to London, he will meet with Matthew.
[If John1 goes to London]4 will f4 [he1 meet Matthew]
Will do better
At a semantic level, conditionals are interpreted via a rule that instructs us to
perform assignment shift, mapping the modal base variable to a set of worlds
determined by intersecting the old modal base with the proposition expressed by
the antecedent. Formally:
¹(if A)(M ODAL f B)ºw,s,g = ¹M ODAL f Bºw,s,g[ f →F∩A]
(recall that F = g( f ) and A = {w : ¹Aºw,s,g = 1})
To illustrate this, consider again (16). Given modal base F, let g ∗ be the assignment that coincides with g except at index 4, which is mapped to the set of worlds
in F at which John goes to London (i.e. g ∗ = g[4 → F∩ ¹John goes to Londonº]).
Then we predict:
¹(16)ºw,s,g = ¹will f4 [he1 meet Matthew]ºw,s,g
Informally, and simplifying, the resulting truth conditions of (16) are:
¹(16)ºw,s,g = true iff John meets Matthew at v where v is the world that
is selected when s is given as input the set of the historical alternatives
(to w) where John goes to London.
Notice that the selected world need not coincide with the actual world or with
the world of evaluation. In particular, for any w such that John does not go to
London at w, then the world selected by s taking w as input must be different from
w itself. In this case, our Stalnakerian semantics and the Ockhamist tradition
This treatment of conditionals also yields a straightforward account of modal
subordination. Consider again Klecha’s example:
If the supplies arrive tomorrow, it will be late in the day. They will contain
three boxes of cereal.
We can predict the relevant interpretation of (5) by assuming that the modal
base variables associated to the two occurrences of will are coindexed:
If the supplies arrive tomorrow, it will f3 be late in the day. They will f3
contain three boxes of cereal.
The antecedent if the supplies arrive tomorrow shifts the value of the relevant
index. But, given coindexing, both occurrences of will are interpreted in the
scope of the relevant supposition.15
The will-would connection
Our treatment of will also allows us to vindicate the morphological connection
between will and would. The precise nature of this connection depends on one’s
views about the meaning of would.
On the one hand, if one assumes a Stalnakerian semantics for would, then the
connection is immediately vindicated: will and would turn out to have exactly the
same meaning—modulo differences in what possibilities are in the modal base in
the two cases. Of course, Stalnaker’s semantics for would, and in particular the
principle of Conditional Excluded Middle that it entails, are controversial. But,
first, notice that all the arguments that we gave above for the scopelessness of
will carry over to the case of would. Moreover, the literature has provided plenty
of further arguments in support of Conditional Excluded Middle (see, e.g., von
Fintel & Iatridou 2002, Williams 2010, Klinedinst 2011).
On the other, the connection is not straightforward if one adopts a Lewisian
semantics for would. In this case, one will have to explain why will deploys a
selection function while would has universal quantificational force. While this is
a nontrivial task, a selection function account of will is better placed at fulfilling
it than an old-fashioned Ockhamist semantics.
Epistemic readings of will
Recall the example we used to introduce epistemic readings of will:
John will be in London by now.
will in (3) is not used to talk about the future, but rather has an epistemic reading.
Predicting how and when will receives an epistemic reading goes beyond the scope
of this paper. These questions connect to general phenomena in the semantics of
modality, and its relationship with tense and aspect, that we can’t cover here.16
To see this, notice that the availability of epistemic readings for will patterns with
the availability of epistemic readings with other modals, like must and have to.
15 We assume that the effects of the shift operated by the conditional antecedent extend beyond the
boundaries of individual sentences. An assumption of this sort seems required by any account of
modal subordination.
16 For some relevant discussion, see Condoravdi (2002), Condoravdi & Deo (2008).
Will do better
John will be in London by now. (Øepistemic/ #historical)
John will go to London tomorrow. (#epistemic/ Øhistorical)
John must/has to be in London by now. (Øepistemic/ #deontic)
John must/has to go to London tomorrow. (#epistemic/ Ødeontic)
The natural way to account for contrasts like the ones in (22-a-b) and (23-a-b) is
to assume that the different prejacents somehow force a different choice of modal
bases for the modals. For example, Condoravdi (2002) suggests that modalized
claims whose prejacent has a reference time in the present (like (22-a) and
(23-a)) rule out non-epistemic modal bases via a constraint requiring that the
modal base be sufficiently diverse.17
Even if we can’t provide an account of the data here, it should be clear that
our semantics provides the tools to build such an account. Once again, the key
move is endowing will with a modal base, contrary to Ockhamist semantics and
in agreement with modal semantics. Once we make this move, we can just draw
on independent work on the interaction between tense, aspect, and modal bases.
Truth, validity, and indeterminacy
§5 offers a compositional semantic account of will, but does not fix the truth
conditions of will-claims. To get the latter, we must define a notion of truth at a
context (this is the stage of the theory that MacFarlane 2003 calls ‘postsemantics’).
Adopting a definition of truth at a context requires us to take sides in the
debate about the indeterminacy of statements about the future. To make the
presentation more concrete, we adopt a specific account of how indeterminacy
affects the semantics, i.e. the version of supervaluationism defended by Barnes
& Cameron (2009) (see also Iacona 2014, who sketches an Ockhamist picture
whose treatment of indeterminacy is parallel to ours). According to this form
of supervaluationism, each context determines a single actual world, but it is
indeterminate which context the utterance takes place in.
In a standard contextualist framework, built on Kaplan (1989), truth at a
context is defined by fixing the values of index parameters to the coordinates of
the context (following Lewis 1980, we take contexts to be concrete situations of
utterance). Here is a formal definition:
Truth at a Context. A is true as uttered at c iff ¹Aºwc ,sc ,gc = 1
17 This is the constraint endorsed by Condoravdi, roughly stated:
Diversity condition. If ðMODAL Añ has a non-epistemic modal base M, then
there are worlds w and v in M such that A is true at w and false at v.
In a slogan: S is true at a context c just in case S is true at the circumstances
fixed by c. Traditional supervaluationist accounts reject Kaplan’s definition of
truth at a context. They maintain that a context of utterance does not fix which
world is to count as actual. These accounts replace truth at a context with a new
definition that allows for sentences that are neither true nor false at the context
of utterance.18 Unlike traditional supervaluationists, we endorse the simple
definition of truth at a context above. Hence, on our account, every sentence is
either true or false at a context.
How, then, can we satisfy the theorists who maintain that the future is
genuinely open? We assume that if the future is open, it is indeterminate which
context the utterance takes place in, and hence it is indeterminate which truth
value the sentence has. An important consequence of this hypothesis is that both
the defender and the opponent of Openness are able to help themselves to our
framework, including the definition of truth (and the two notions of validity that
we give below). Their disagreement is moved out of the semantic apparatus
entirely, as the defender of Openness denies, and the opponent of Openness
claims, that a concrete situation of utterance determines a unique context.
The notion of truth at a context is important for a number of reasons. One of
them is that (following Kaplan himself) we can use it to define a plausible notion
of validity. On this notion, an argument is valid just in case it preserves truth at
a context: no context makes the premises true and the conclusion false.
Validity1 . A1 ,..., An 1 B iff, for any context c such that A1 ,..., An
are true at c, B is also true at c.
We should flag an important consequence of our moving the indeterminacy
outside of the semantics: unlike traditional supervaluationism, our logic for the
relevant fragment of the language is allowed to be straightforwardly classical. It
will also be helpful to appeal to a second notion of validity, one that captures
preservation of truth at a point of evaluation.
Validity2 . A1 ,..., An 2 B iff, for any triple 〈w,s, g〉 such that ¹A1 ºw,s,g =
1, ..., ¹An ºw,s,g = 1, ¹Bºw,s,g = 1
A single sentence A is valid1 just in case 1 A; similarly for valid2 . Note that
every argument that is valid2 is valid1 , but not viceversa. When A and B are such
that A 1 B and B 1 A, we say that they are equivalent1 (similarly for equivalent2 ). Here too we have that every equivalent2 pair is equivalent1 (though
18 For completeness, here is a sample definition of truth at a context that fits the traditional
supervaluationist’s desiderata: A is true as uttered at c iff for all worlds v that are historically
possible in c , ¹Aº v,sc ,gc = 1.
Will do better
not vice-versa). These connections between these logical notions are important.
Although validity1 (and equivalence1 ) are more significant notions in the overall
architecture of the theory, we can establish them by way of establishing validity2
(and equivalence2 ).
Consequences for the logic of will
Having acquired a notion of validity, we are ready to explore the main consequences of our apparatus. We start by vindicating our second desideratum, i.e. the
scopelessness of will. First, we notice that the semantics satisfies the excluded
middle property:
Will Excluded Middle. ðwill f A ∨ will f not Añ is valid2 (hence
valid1 ).
We use ‘∨’ for disjunction to highlight that the sentence is valid on the assumption that or is boolean disjunction.19 The argument for excluded middle also
establishes a related fact:
Negation Swap. ðwill f not Añ and ðnot will f Añ are equivalent2
(hence equivalent1 )
This explains why we do not perceive different scopes for negation despite the
fact that will is a modal.
The analysis also entails that will-conditionals satisfy a principle of conditional
excluded middle. Interestingly, this holds whether or not conditionals in general
satisfy this principle.
Compositional CEM for will-Conditionals. For any point 〈w,s, g〉,
¹(if B)(will f A)ºw,s,g = 1 or ¹(if B)(will f not A)ºw,s,g = 1
Note that this and all the following about conditionals are restricted to the case
in which the conditional antecedent is compatible with the modal base.
Since Compositional CEM holds at any point of evaluation,20 we get:
19 PROOF: let 〈w,s, g〉 be an arbitrary point of evaluation. We have that ðwill f A ∨ will f not Añ is
true at 〈w,s, g〉 iff either A is true at s(w,F) or false at s(w,F). But the right-hand side of the
biconditional is always true (since, for any set of worlds S and any world w, it is always the case
that w either belongs or doesn’t belong to it). Hence, ðwill f A ∨ will f not Añ is true at 〈w,s, g〉,
hence ðwill f A ∨ will f not Añ is true at any point of evaluation.
20 Whether ð(if B)(will f A)ñ is true at 〈w,s, g〉 boils down to the truth of A at 〈w,s, g[ f 7→ F ∩ B]〉.
If it is not true, it must be false, but in that case, ð(if B)(will f not A)ñ must be true.
Postsemantic CEM for will-Conditionals.
ð(if B)(will f A) ∨ (if B)(will f not A)ñ is valid2 (hence valid1 )
Relatedly, there is only one way of negating the consequents of conditionals.
Narrow Negation Swap in Conditionals. ð(if B)(will f not A)ñ
and ð(if B)(not will f A)ñ are equivalent2 (and hence equivalent1 ).
This is a trivial consequence of the non-conditional negation swap, because (if
B) merely operates on g. More importantly, we can derive:
Wide Negation Swap in Conditionals. ð(if B)(will f not A)ñ and
ðnot (if B)(will f A)ñ are equivalent2 (and hence equivalent1 ).21
This completes our illustration of the basic logical implications of our semantics
as well as our explanation of how modal theorists can explain the scopelesness
of will.
Belief and doubt in will-claims
Probabilities of simple will-claims
Recall the cognitive problem from §2.3. Ordinary agents are uncertain about
the future. On one natural way to understand this uncertainty, this means that
ordinary agents have nonextreme degrees of belief in the propositions expressed
by will-claims. Moreover, at least in some cases, it seems that this uncertainty is
rationally permissible, if not rationally required. An adequate theory of will should
vindicate this intuition.
This problem should be distinguished from a different, important problem
surrounding belief in future claims. Suppose that the future is objectively open,
in the sense defined above. In particular, suppose that branching theorists in the
style of Thomason (1970), Belnap & Green (1994), Belnap et al. (2001) are right
about the metaphysics of branching: there are several possible futures, each of
which share the segment that we occupy at the present time. In this case, it is
unclear what we mean when we say that the probability of an open proposition—
say, the proposition that Cynthia will wear a Warriors cap tomorrow—is r. If
it is genuinely open whether Cynthia will wear a Warriors cap, then there are
(at least) two ‘equally real’ futures. In one of them she wears the cap, while
21 PROOF: ¹not (if B)(will f A)ºw,s,g = 1 iff ¹(if B)(will f A)ºw,s,g = 0 iff ¹will f Aºw,s,g[ f 7→F∩B] =0
iff ¹not will f Aºw,s,g[ f 7→F∩B] =1 iff ¹(if B)(not will f A)ºw,s,g =1. Together with Narrow Negation
Swap, this equivalence entails Wide Negation Swap.
Will do better
in the other she does not. This problem has received attention in philosophy
of physics (in particular, by defenders of the Everett interpretation of quantum
mechanics—see, among others, Wallace 2014). We are not contributing to this
project here. Rather, we are taking for granted that it somehow makes sense
to assign nonextreme credences to propositions about the future (and hence,
derivatively, to the claims that express them).22
Let us then return to our Sports Fan example. Recall: every day, Cynthia
tosses a fair die and, on the basis of the outcome, decides whether to wear a
Giants hat, a Warriors hat, or no hat. Consider a rational agent who assigns
credence 1/3 to each of the three possibilities. Against this background, what
we want to show is that the proposition expressed by (10) (repeated below) on
our account also gets credence 1/3.
Cynthia will wear a Warriors cap tomorrow
Before we start, let us remind you that this is a nontrivial task. In fact, as we
pointed out in §2.4, contemporary accounts in the linguistic semantics tradition
fail this task. The reason is that these accounts declare ðIt will be the case that Añ
true just in case A is true at every possible future. This semantics makes (10)
false, and hence predicts that a rational agent should assign credence zero to the
proposition it expresses.
Let us spell out some basic assumptions. In keeping with standard possible
worlds semantics for attitudes, assume that credences are defined over sets of
worlds. In particular, assume that an agent’s credences at a given time may be
modeled by a probability function µ satisfying the usual constraints. For example,
let µ model an agent’s credences at the current point in time. Let µ(A) = 1/3,
where A is the set of worlds where Cynthia wears a Warriors cap. Our task
is to check that µ(PROPW ) = 1/3, where PROPW is the content our semantics
associates to an utterance of (10).
For current purposes, we can take the content expressed by the utterance of
a will-sentence at a given context to be the set of worlds such that the utterance
is true as evaluated at those worlds. Formally:
Content of A at c: kAkc = {w : ¹Aºw,sc ,g c = 1}
In what follows, we suppress reference to the context to avoid clutter.
22 If you’re skeptical about these claims, you may take our arguments in this section as providing
a good litmus test for our semantics of will. What we’re going to show is that if our theories
of credences warrants assigning the attitudes that seem intuitive offhand, then the semantics
delivers contents that vindicate the intuitive assignment of credences.
it is easy to see that this yields exactly the verdict we need. Take our example:
kCynthia will wear a Warriors capk is just the set of worlds in which Cynthia
wears a Warriors cap. On the assumption that the credence that our agent assigns
to Warriors-cap-worlds is 1/3, she will also assign credence 1/3 to the proposition
expressed by (10). More generally, we obtain the result that, for any prejacent
A, kwill f Ak = kAk.
Probabilities of complex will-sentences
The probabilities of Boolean compounds of will-sentences work out as one would
intuitively expect. In virtue of our earlier result that the scope of negation does
not matter, we have that:
knot will f Ak = {w : w is an ¬A-world} = k¬Ak
In our example, kIt is not the case that Cynthia will wear a Giants capk = kCynthia
will not wear a Giants capk. Both propositions have probability 2/3 (according
to µ). Similar results holds for conjunction and disjunction, so that: kwill (A &
B)k = kwill Ak ∩ kwill Bk and kwill (A ∨ B)k = kwill Ak ∪ kwill Bk.
The case of conditionals is more complex. It is well-known that standard possible worlds semantics for conditionals fails to vindicate, in general, the intuitive
assignments of probabilities to conditional sentences.23 Since our account of
will mirrors one of the standard semantic proposals for conditionals, it shares
this feature.
It is helpful to give a sense of how standard semantics for conditionals fails to
vindicate intuitive probability judgments. Consider again the Sports Fan scenario,
and take the conditional:
If Cynthia wears a cap, she will wear a Warriors cap.
Recall that Cynthia decides to wear a Warriors cap, a Giants cap, or no cap,
depending on a random process that makes each of the three options 1/3 likely.
Accordingly, suppose that (24) is evaluated against the toy modal base we described in section 3, and consisting of a Warriors-cap-world, a Giants-cap-world,
and a no-cap-world. It seems natural to say that your degree of belief in (24)
should be (or, at the very least, may be) 1/2. But our semantics can’t vindicate
this result.
23 See Lewis (1979b), Lewis (1986b), Hájek & Hall (1994). It is often assumed that the intuitive
probabilities of conditional sentences should match the conditional probabilities of the consequent,
given the antecedent. But one doesn’t need to endorse this general thesis (which has been called
into question; see e.g. Kaufmann (2004)) to see the problem. All we need is that there are
examples in which such an assignment is plausible, such as (24) in the main text.
Will do better
To see this, consider the content of (24):
kIf cap, will f Warriors capk =
{w: ¹If cap, will f Warriors capºw,s,g = 1} =
{w: ¹will f Warriors capºw,s,g[ f →F∩cap] = 1} =
{w: ¹Warriors capºs(w,F∩cap),s,g[ f →F∩cap] = 1}
A world belongs to this proposition if the selection function maps to it when
given as input that world and the restricted modal base. it is easy to show that
the proposition in (25) cannot have probability 1/2. The basic point is that,
given that in our model we have only three worlds, each of which has probability
1/3, no proposition (i.e, no set of worlds) can have probability 1/2.24
One could respond that the problem depends on the fact that we have used a
modal base that is too simple. If we add enough worlds to the modal base, we
will be able to assign to (25) a content that gets probability 1/2. This is correct,
but the point illustrated by our toy example will still hold for some conditional
or other. Alan Hájek (1989, 2012) has pointed out that for any (nontrivial and
finite-ranged) probability function P r, the conditional probability values assigned
on the basis of P r outnumber the unconditional probability values assigned by
P r. Hence there will always be some conditional probability value that doesn’t
find a match in the probability of any proposition—exactly as it happens in our
example, where no proposition has probability 1/2.
Before closing, let us gesture towards a way of refining our ideas that will
yield better results for conditionals. We build on a line of thinking of conditionals
that has been developed over the past three decades (McGee 1989, Jeffrey &
Stalnaker 1994, Bradley 2012). So far, we have assumed that the credences of a
rational agent are distributed over an algebra of possible world propositions. But
there is a second dimension of uncertainty, which this model doesn’t capture: an
agent may be uncertain about which world is selected by the selection function.
Go back to our example and consider the no-cap-world z. What is the value of
the selection function, when the relevant arguments are z and the proposition
that Cynthia wears a cap? Both the answers ‘the Warriors-cap-world w’ and ‘the
Giants-cap-world v’ seem to be open epistemic possibilities. Being uncertain
between these two answers, of course, would mean being uncertain about which
24 PROOF: The modal base contains three worlds: the Warriors-cap-world w, the Giants-cap-world
v, and the no-cap-world z; each has probability 1/3. w is a member of k(24)k and v is not. The
question is whether z is. This depends on the value of the selection function when z is the input
world. If s(z,F ∩ cap) is w, then z is also a member of k(24)k; otherwise, not. In the former case,
the probability of (24) is 2/3, and in the latter 1/3. Either way, that probability is different from
between two candidates for the selection function (call them ‘sw ’ and ‘s v ’) is the
‘correct’ one.
To register this kind of uncertainty alongside uncertainty about which world
is actual, we need to refine the elements of the underlying algebra. Rather than
worlds, we may use pairs of consisting of a world and a selection function. This
also involves modifying our notion of content: we need to take contents to be
sets of pairs of a world and a selection function.25
0 0
2D content of A at c: 〈〈A〉〉 = {〈w0 ,s0 〉 : ¹Aºw ,s ,g c = 1}
It is straightforward to see how, on the new picture, (24) may be assigned
probability 1/2. We now have an algebra of six possibilities, consisting of the
{〈w,sw 〉,〈v,sw 〉,〈z,sw 〉,〈w,s v 〉,〈v,s v 〉,〈z,s v 〉}
it is easy to show that probability distribution that assigns to each of these pairs
probability 1/6 assigns probability 1/2 to (24).
Traditionally, will has been treated as a tense in philosophy, and as a modal in
linguistics. Linguists are right: there is conclusive evidence that will is a modal.
At the same time, all existing modal theories fail to deliver some important
desiderata. In particular, they cannot be integrated with our intuitive attitudes
towards the future. We have suggested that will is indeed a modal, but doesn’t
have a quantificational semantics. Rather, will selects the ‘one actual future’ out
of the set of historical alternatives at the time of utterance. Besides validating
the evidence for the modal character of will, this account predicts a range of
important logical interactions for will and dovetails well with intuitions about
the cognitive role of future statements—thus doing better than any other theory
on the market.
25 See Bradley (2012) for a much more extensive development of the idea.
Will do better
Abusch, Dorit. 1997. Sequence of tense and temporal de re. Linguistics and
Philosophy 20(1). 1–50.
Abusch, Dorit. 1998. Generalizing tense semantics for future contexts. In Rothstein S. (ed.), Events and Grammar (volume 70 of Studies in Linguistics and
Philosophy, 13–33. Kluwer.
Barnes, Elizabeth & Ross Cameron. 2009. The open future: bivalence, determinism and ontology. Philosophical Studies 146. 291–309.
Belnap, Nuel & Mitchell Green. 1994. Indeterminism and the thin red line.
Philosophical perspectives 3. 365–388.
Belnap, Nuel, Michael Perloff & Ming Xu. 2001. Facing the Future. Oxford
University Press.
Bradley, Richard. 2012. Multidimensional possible-world semantics for conditionals. Philosophical Review 121(4). 539–571.
Cariani, Fabrizio. unpublished. Predictions and modality. Manuscript, Northwestern University.
Condoravdi, Cleo. 2002. Temporal interpretation of modals: Modals for the
present and for the past. In D. Beaver, L. Casillas, B. Clark & S. Kaufmann
(eds.), The Construction of Meaning, Palo Alto, CA: CSLI Publications.
Condoravdi, Cleo & Ashwini Deo. 2008. Aspect shifts in indo-aryan. In Proceedings
of the 18th International Congress of Linguistics, Seoul, Korea, 1–20.
Copley, Bridget. 2009. The Semantics of the Future. Routledge.
Edgington, Dorothy. 2008. Counterfactuals. Proceedings of the Aristotelian Society
108. 1–21.
Enç, Mürvet. 1996. Tense and modality. In Shalom Lappin (ed.), The Handbook
of Contemporary Semantic Theory, 345–358. Blackwell.
von Fintel, Kai. 1994. Restrictions on Quantifier Domains.: UMass, Amherst
von Fintel, Kai. 1997. Bare plurals, bare conditionals and Only. Journal of
Semantics 14(1).
von Fintel, Kai & Irene Heim. 2011. Notes on intensional semantics. MIT.
von Fintel, Kai & Sabine Iatridou. 2002. If and when if -clauses can restrict
quantifiers. MIT.
Hájek, Alan. 1989. Probabilities of conditionals: revisited. Journal of Philosophical
Logic 18(4). 423–428.
Hájek, Alan. 2012. The fall of “Adams’ thesis”. Journal of Logic, Language and
Information 21(2). 145–161.
Hájek, Alan & Ned Hall. 1994. The hypothesis of the conditional construial of
conditional probability. In Ellery Eells & Brian Skyrms (eds.), Probability
and conditionals: Belief revision and rational decision, 75–113. Cambridge
University Press.
Heim, Irene & Angelika Kratzer. 1998. Semantics in Generative Grammar. Blackwell.
Hiddleston, Eric. 2005. A causal theory of counterfactuals. Noûs 39(4). 632–657.
Iacona, Andrea. 2014. Ockhamism without the thin red line. Synthese 191.
Jeffrey, Richard & Robert Stalnaker. 1994. Conditionals as random variables,
75–111. Cambridge University Press, Cambridge.
Kaplan, David. 1989. Demonstratives. In J. Almog, J. Perry & H. Wettstein (eds.),
Themes from Kaplan, Oxford University Press.
Kaufmann, Stefan. 2004. Conditioning against the grain. Journal of Philosophical
Logic 33(6). 583–606.
Kaufmann, Stefan. 2005. Conditional truth and future reference. Journal of
Semantics 22(3). 231–280.
Klecha, Peter. 2013. Diagnosing modality in predictive expressions. Journal of
Semantics 31(3). 443–455.
Klinedinst, Nathan. 2011. Quantified conditionals and conditional excluded
middle. Journal of Semantics 28(1). 149–170.
Korzukhin, Theodore. 2014. Dominance conditionals and the newcomb problem.
Philosophers’ Imprint 14(9). 1–20.
Kratzer, Angelika. 1977. What ‘must’ and ‘can’ must and can mean. Linguistics
and Philosophy 1(3). 337–355.
Kratzer, Angelika. 1981a. The notional category of modality. In B. Partee &
P. Portner (eds.), Formal Semantics: the Essential Readings, Blackwell.
Kratzer, Angelika. 1981b. Partition and revision: The semantics of counterfactuals. Journal of Philosophical Logic 10(2). 201–216.
Kratzer, Angelika. 1991a. Conditionals. In A. von Stechow & D. Wunderlich
(ed.), Semantics: An International Handbook of Contemporary Research, De
Gruyter. From the Semantics Archive.
Kratzer, Angelika. 1991b. Modality. In A. von Stechow & D. Wunderlich (ed.),
Semantics: An International Handbook of Contemporary Research, De Gruyter.
Kratzer, Angelika. 2012. Modals and Conditionals. Oxford University Press.
Lewis, David K. 1975. Adverbs of quantification. In Edward L. Keenan (ed.),
Formal semantics of natural language, Cambridge University Press.
Lewis, David K. 1979a. Counterfactual dependence and time’s arrow. Noûs 13(4).
Lewis, David K. 1979b. Probabilities of Conditionals and Conditional Probabilities.
The Philosophical Review 84. 297–315.
Will do better
Lewis, David K. 1980. Index, context and content. In S. Kanger & Sven Öhman
(eds.), Philosophy and Grammar, Reidel.
Lewis, David K. 1986a. On the Plurality of Worlds. Blackwell.
Lewis, David K. 1986b. Probabilities of conditionals and conditional probabilities
II. The Philosophical Review 95. 581–589.
Lewis, David K. 1986c. A subjectivist’s guide to objective chance. In Philosophical
Papers, Vol. II, 263–293. Oxford University Press.
MacFarlane, John. 2003. Future contingents and relative truth. Philosophical
Quarterly 53. 321–336.
MacFarlane, John. 2008. Truth in the garden of forking paths. In M. Kölbel &
M. Garcia-Carpintero (eds.), Relative Truth, Oxford: Oxford University Press.
MacFarlane, John. 2014. Assessment Sensitivity. Oxford University Press.
McGee, Vann. 1989. Conditional probabilities and compounds of conditionals.
The Philosophical Review 485–541.
Moss, Sarah. 2014. Subjunctive credences and semantic humility. Philosophy and
Phenomenological Research 87. 251–278.
Palmer, F. R. 1987. Mood and Modality. Cambridge University Press.
Prior, Arthur. 1967. Past, Present and Future. Oxford University Press.
Roberts, Craige. 1989. Modal subordination and pronominal anaphora in discourse. Linguistics and philosophy 12(6). 683–721.
Schulz, Moritz. 2014. Counterfactuals and arbitrariness. Mind 123(492). 1021–
Stalnaker, Robert C. 1968. A theory of conditionals. Studies in Logical Theory,
American Philosophical Quarterly Monograph Series 2. 98–112.
Thomason, Richmond H. 1970. Indeterminist time and truth-value gaps. Theoria
18(3). 264–281.
Wallace, David. 2014. The Emergent Multiverse. Oxford University Press.
Williams, Robert J. 2010. Defending conditional excluded middle. Noûs 44.